The SMS Sunset: What Encrypted RCS Means for the One-Time Passcode

In May 2026, Apple and Google began rolling out end-to-end encryption for RCS messages between iPhone and Android, enabled by default and expanding over time. It is a significant, hard-won privacy improvement for ordinary text messaging. It is also, and this needs saying plainly, almost entirely irrelevant to the security of the one-time passcode — which remains the part of multi-factor authentication that organizations most need to retire.

Encryption Protects the Message, Not the Number

The confusion is understandable. If your messages are end-to-end encrypted, surely your login codes are safe? They are not, because the weakness of SMS-based authentication was never the plaintext of the message. The weakness is that the phone number itself is the credential, and the number is controlled by people you have no relationship with. A SIM swap, a carrier port-out, or a compromised roaming path lets an attacker receive the code regardless of whether the transport is encrypted. End-to-end encryption hardens the pipe; it does not harden the identity binding at the end of it. Recent rollouts also remain partial — group conversations, some carrier routes, and business messaging paths do not uniformly inherit the same end-to-end guarantees, and codes delivered through business messaging channels terminate at the messaging provider rather than the device.

Why the Industry Kept Warning About This

Published authentication guidance has for years placed short-message-based one-time codes in a restricted category, permitted but discouraged, precisely because the underlying transport and number control offer no strong guarantees. The additional exposure is mundane and often overlooked: codes visible in previews on locked screens, malware with SMS read permissions, message forwarding, and operator-side access. None of those are fixed by encrypting the message body between two clients.

A Practical Retirement Plan

• Inventory the dependency: Find every place SMS codes are still used, and rank by consequence. Privileged accounts, remote access, and financial approvals get fixed first.
• Replace with phishing-resistant factors: Passkeys and hardware security keys are the destination for high-value access. Application-based push with number matching is an acceptable intermediate that removes the number from the trust path.
• Harden what remains: Where SMS must stay as a fallback, shorten code lifetimes, enforce single use and strict rate limits, suppress lock-screen previews, and encourage carrier port-out locks and SIM PINs on personal lines.
• Watch the recovery path: In practice, the SMS fallback survives as the reset mechanism long after the primary login has moved on — which is exactly where attackers will aim.

There is a fair criticism of this position: SMS codes are far better than no second factor, and for a small organization with no identity budget they are still a meaningful improvement over a password alone. That argument is correct as a starting point and wrong as a destination. The right reading of the May 2026 RCS rollout is that ordinary messaging got meaningfully more private while authentication quietly confirmed how little it ever depended on that privacy.